The document discusses using bacterial source tracking (BST) to identify sources of E. coli contamination in watersheds. It describes how Texas developed a BST library with over 1,600 E. coli isolates from different human and animal sources. Studies using the library found that wildlife sources accounted for about 50% of isolates in rural watersheds on average, while correlations between land use and sources were limited. The document concludes by discussing future areas of research like assessing urban watersheds and improving library-independent BST methods.

1. Improving Watershed Planning
Using Bacterial Source
Tracking
2015 SWCS Conference
K. Wagner, G. DiGiovanni, E. Casarez, J.
Truesdale, P. Wanjugi, T. Gentry, L. Gregory

2. Bacteria/Pathogens
The #1 Cause of Water Quality Impairment in Texas
Rank Rivers/Streams Lakes/Reservoirs Bays/Estuaries
1 Pathogens (16%) Mercury (43%) Mercury (33%)
2 Sediment (12%) Nutrients (18%) PCBs (23%)
3 Nutrients (10%) PCBs (16%) Pathogens (21%)
4
Organic enrichment /
Oxygen Depletion (9%)
Turbidity (8%)
Organic enrichment /
Oxygen Depletion (17%)
5 PCBs (8%)
Organic enrichment /
Oxygen Depletion (8%)
Dioxins (14%)
The #1 Cause of River/Stream Impairment in U.S.

3. Where did the Bacteria (E. coli) Come From?
• Potential sources
• Humans
• Domesticated animals
• Wildlife
• Methods for determining sources
• Source survey
• Modeling
• Bacterial source tracking (BST)

4. PREMISE BEHIND BST
Different guts  Different adaptations
 Different E. coli strains 
Genetic Differences
Phenotypic Differences

5. Classifications of BST Methods

6. Establishment of Texas BST
Program (2007)
• Two DNA fingerprinting methods selected:
• Enterobacterial repetitive intergenic
consensus sequence-polymerase chain
reaction (ERIC-PCR)
• RiboPrinting® (RP)
• Required BST Library Development

7. Development of Texas
E. coli BST Library
Sources
Isolate
E. coli
DNA
Fingerprint
Add to
Library

8. Texas E. coli BST Library
• Contains
• 1,669 E. coli isolates
• From 1,455 different
fecal samples
• Representing >50
animal subclasses
• Collected from 13
watersheds (& growing)
across Texas
Wildlife
41%
Domestic
Animals
34%
Human
25%

9. Use of Texas E. coli BST Library for
Identifying Water Isolates
Isolate
E. coli
DNA
Fingerprint
Compare
to Library
Source
ID

10. Texas E. coli BST Library composition
& rates of correct classification (RCC)
Source Class
Number of
Isolates
Number of
Samples
Library
Composition and
Expected Random
Rate of Correct
Classification
Calculated Rate of
Correct
Classification
(RCC)
RCC to Random
Ratio***
Left Unidentified
(unique patterns)
HUMAN 364 315 24% 100 4.2 22
DOMESTIC
ANIMALS
531 474 35% 100 2.9 19
Pets 86 76 6% 83 13.8 40
Cattle 237 207 16% 93 5.8 11
Avian Livestock 96 83 6% 89 14.8 25
Other Non-Avian
Livestock
112 108 7% 90 12.9 14
WILDLIFE 629 569 41% 100 2.4 19
Avian Wildlife 239 221 16% 85 5.3 21
Non-Avian Wildlife 390 348 26% 92 3.5 17
Overall 1524 1358
3-way = 100%
7-way = 92%
20%

11. Texas BST Studies To Date
Typical Landuse in 11 BST
Watersheds

12. Wildlife
51%
Human
10%
Domestic
Animals
27%
Unidentified
12%
3-Way Split
(averages based on findings in 11 watersheds)
Non-Avian
Wildlife
32%
Avian
Wildlife
18%
Pets
5%
All Livestock
24%Human
10%
Unidentified
11%
5-Way Split
(averages based on findings in 10 watersheds)
Non-Avian
Wildlife
32%
Avian
Wildlife
18%
Pets
5%
Other Non-
Avian
Livestock
5%
Avian
Livestock
5%
Cattle
13%
Human
10%
Unidentified
12%
7-Way Split
(averages based on findings in 7 watersheds)

13. Relation of Landuse to BST Results
Developed vs Pet & Human Contributions
• Significant correlation between % of watershed
developed and % of isolates from pets
• No correlation between % of watershed
developed and % of isolates from human
R² = 0.5767
0%
2%
4%
6%
8%
10%
12%
14%
16%
18%
0% 5% 10% 15% 20% 25% 30%
%ofisolatesfrompets
% of watershed developed
R² = 0.1133
0%
2%
4%
6%
8%
10%
12%
14%
16%
18%
0% 5% 10% 15% 20% 25% 30%
%ofisolatesfromhuman
% of watershed developed

14. Relation of Landuse to BST Results
Cattle
• No correlation
between watershed
landuse and % of
isolates from cattle
R² = 0.4049
0%
5%
10%
15%
20%
25%
0% 20% 40% 60% 80%
%ofisolatescattle
% of watershed pasture/range
R² = 2E-06
0%
5%
10%
15%
20%
25%
0% 20% 40% 60% 80%
%ofisolatescattle
% of watershed pasture
R² = 0.2083
0%
5%
10%
15%
20%
25%
0% 20% 40% 60% 80%
%ofisolatescattle
% of watershed range

15. Relation of Landuse to BST Results
Wildlife
• Only one significant
correlation observed:
– Btwn % of watershed as
pasture/range/forest & % of
isolates as non-avian wildlife
R² = 0.0067
0%
10%
20%
30%
40%
50%
60%
70%
80% 85% 90% 95% 100%
%ofisolateswildlife
% of watershed pasture/forest/range
R² = 0.4986
0%
10%
20%
30%
40%
50%
60%
70%
80% 85% 90% 95% 100%
%ofisolatesnon-avianwildlife
% of watershed pasture/forest/range
R² = 0.2073
0%
10%
20%
30%
40%
50%
60%
70%
80% 85% 90% 95% 100%
%ofisolatesavianwildlife
% of watershed pasture/forest/range

16. Conclusions
• BST performing well & tremendously helpful
in identifying significant bacteria sources
• Wildlife is source of 50% of isolates in
predominately rural watersheds
• Generally no correlations between landuse
and isolate source (i.e. LULC may not be
good predictor of bacteria sources)

17. Future Methods & Approaches
1. Assess urban watersheds
2. Identify the “Unidentified”
– Continue expansion of BST library
– Evaluate other sources of E. coli

18. Future Methods & Approaches
3. Improve Library Independent BST (Bacteroidales)
– Genotypic detection of microorganisms based on marker genes
– Does not require known-source library
– Rapid & less expensive than library methods
Extract
DNA
PCR amplify
target sequence
+ + - -
Cycle
5 10 15 20 25 30 35 40 45 50 55
Norm.Fluoro.
1.4
1.3
1.2
1.1
1.0
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0.0 Threshold
1 2 43

19. Questions?
• Kevin Wagner
• TWRI Assoc. Director
• 979-845-2649
• klwagner@ag.tamu.edu
• George Di Giovanni
• Professor, UT School of
Public Health – El Paso
• 915-747-8509
• george.d.digiovanni@uth.tmc.edu
• Terry Gentry
• Assoc. Professor, Texas
A&M AgriLife Research
• 979-845-5323
• tgentry@ag.tamu.edu